Estrogen receptor (ER) breast cancers constitute nearly 75% of all cases and half of all deaths from the disease each year. These tumors are termed luminal and are remarkably heterogeneous in their expression of ER. Tumors that contain only a small percent of ER immunoreactive cells are candidates for treatment with endocrine therapies. However, nearly 40% of luminal tumors will become endocrine resistant and relapse, which persistently occurs over a 5-10 year period. The role of various ER?subpopulations within ER+ tumors is largely unknown. In our previous work, we identified that half of luminal tumors contain an ER? and progesterone receptor (PR)? subpopulation marked by cytokeratin 5 (CK5), a normal breast stem cell marker, and indicator of poor prognosis in breast cancer. CK5+ER?PR? cells display cancer stem cell (CSC) properties such as resistance to endocrine and chemotherapy, and enhanced mammosphere and tumor initiation capacity. We originally described that the natural hormone progesterone and progestins, synthetic analogs used for menopausal hormone therapies, expand the CK5+ breast cancer cell population. We have confirmed this using our original collection of luminal ER+PR+ breast cancer patient-derived xenografts (PDX). Furthermore, we have defined that retinoic acid can block or reverse progestin-induced CK5+ cells. The goal of this proposal is to further investigate the hormone and nuclear receptor mediated mechanisms by which CK5+ER?PR? breast cancer cells are regulated, and how they contribute to disease progression. The hypothesis to be tested is that progestins through PR directly regulate the CK5 gene to produce CK5+ER?PR? CSCs that promote endocrine resistance and tumor recurrence. Our approach will utilize luminal breast cancer cell lines with traceable CK5 expression, our luminal breast cancer PDX that retain the natural mixture of CK5?ER+PR+ and CK5+ER?PR? tumor cells, and cohorts of endocrine therapy treated patient clinical samples. Experiments will 1) Determine the mechanism of PR induction of the CK5 gene through an essential progesterone responsive element (PRE); 2) Evaluate how progestins convert single ER+PR+ cells to ER?PR? mammosphere-forming CSCs; and 3) Define if blocking CK5+ cells during endocrine therapy will reduce recurrence. These studies have implications for individual tumor cell plasticity, and could explain how progestins accelerate tumor growth during hormone replacement therapies. They also seek to explain if CK5+ ER?PR? cells contribute to the increased recurrence of some luminal tumors. Our long term goal is to implement strategic and simultaneous management of heterogeneous luminal breast cell populations with the ultimate goal of reducing the incidence of endocrine refractory breast cancer.